Relationship among paretic knee extension strength, maximum weight-bearing, and gait speed in patients with stroke

'Stroke Cerebrovasc Dis 1991;1 :65-69 © 1991 Demos Publications

Relationship Among Paretic Knee Extension Strength, Maximum Weight-Bearing, and Gait Speed in Patients with Stroke Richard W. Bohannon, PT, Ed.D., NCS

The purpose of this study was to determine the intercorrelations among maximum weight-bearing through the paretic lower extremity, paretic knee extension strength, and comfortable gait speed in hemiparetic patients following stroke. Twenty patients, who could ambulate at least 10.0 m with no more assistance than contact guarding of one person, participated. Isometric muscle strength (force) was measured with a hand-held dynamometer. Maximum weight-bearing measurements were normalized against (divided by) body weight. Gait speed was determined using a dig ital stopwatch as subjects walked 8.0 m. Significant correlations (r > 0.611) were demonstrated among the three variables. The relationship between maximum weight-bearing and gait speed was the strongest and was best described by a curvilinear model (R = 0.830). The results suggest that both maximum weight-bearing through the paretic lower extremity and paretic knee extension strength are valid predictors of gait speed among patients with stroke but that the former is superior to the latter for sucha purpose. Given theirsimplicity and objectivity, these measures can be recommended for predicting and documenting improvement in patients with stroke undergoing rehabilitation. Key Words: Stroke-Weight-bearing-Gait speed-Knee strength.

Toprevent collapse during the stance phase of gait, some of the muscles of the supporting lower limb contract to create an extensor moment (torque). The greater the speed of walking, the greater the extensor moment required (1). Although the stance phase extensor moment can be produced by any combination of hip, knee, and ankle extensor muscle activity, the knee extensor muscles normally make a major contribution to the total extensor moment (2). In patients From the School of Allied Health, University of Connecticut, Storrs, and the Department ofRehabilitation, Hartford Hospital, Hartford, CT, U.S.A. Address correspondence and reprint requests to Dr. R W. Bohannon at School of Allied Health, U-I01, University of Connecticut, Storrs, CT 06269, U.S.A.

with stroke, the maximum force or torque (strength) that the knee extensor muscles can produce during isolated testing can be greatly reduced on the paretic side; gait speed tends to be reduced proportionately (3-6). Although different measures of knee extensor muscle strength have been validated as indicators of walking capacity in patients with stroke, the measurements are not without limitations. There are two specific limitations. First, impaired performance of the knee extensor muscles may vary depending on the circumstances of their activation. More specifically, the muscles may function more effectively during weight-bearing than during isolated testing (7) and more effectively during eccentric than during concentric contractions (8). Second, because other muscles contribute to the production of extensor (and

1STROKE CEREBROVASC DIS, VOL. I, NO.2,

1991

65

R. W. BOHANNON

flexor) moments during gait, diminished gait performance among patients with stroke cannot be attributed entirely to a single muscle group such as the knee extensors. Depending on the degree of paresis among other muscle groups and the actual motor strategies employed by the patients with stroke during gait, the demand on the paretic knee extensor muscles may vary. A measure without the two aforementioned limitations is maximum weight-bearing through the paretic lower extremity. Measurements of maximum weight-bearing can reflect the combined capacity of multiple lower extremity muscle groups to create extensor moments and prevent collapse of the paretic lower extremity during eccentric loading. Although numerous reports describe the weight-bearing status of patients with stroke, the descriptions typically describe weight-bearing asymmetries during comfortable stance, not the magnitude of weight-bearing during efforts to maximally weight-bear through the paretic lower extremity (9-13). Granting that the former measurement has been shown to correlate significantly with some measures of gait performance (9), the latter more closely reflects the requirements of gait. In spite of the theoretical advantages of maximum weight-bearing over knee extension strength as a predictor of walking speed, the magnitude of the correlation between maximum weight-bearing and gait speed has not been reported for patients with stroke. Neither has the correlation between maximum weight-bearing and knee extension strength been established. Before impairments such as reduced weight-bearing or decreased muscle strength can be advocated for measurement or as appropriate targets for intervention with patients with stroke, their relationship with disability must be established. Since gait speed is objective and has ramifications for community ambulation (14,15), it is a particularly useful criterion measurement of disability. The purpose of this study was to determine the intercorrelations among maximum weight-bearing through the paretic lower extremity, paretic knee extension strength, and comfortable gait speed. The following hypotheses were tested. (a) There is a positive and significant correlation between maximum weight-bearing through the paretic lower extremity and comfortable gait speed. (b) There is a positive and significant correlation between paretic knee extension strength and comfortable gait speed. (c) There is a positive and Significant correlation between paretic knee extension strength and maximum weight-bearing through the paretic lower extremity.

66

I STROKE CEREBROVASC DIS. VOL

2, NO.2, 1991

Method Subjects Twenty consecutive hemiparetic patients were included in this study after they provided informed consent. They were referred to physical therapy and met five criteria: admitted for their first stroke, could follow three-part verbal instructions, demonstrated no overt unilateral neglect or sensory loss, manifested no other neurologic or orthopedic problems of consequence, and could ambulate at least 10.0 m with no more assistance than contact guarding of one person. Eight of the subjects had right hemiparesis and 12 had left hemiparesis. Ten were men and 10 were women. The mean ± SO and (range) of their ages and times since onset were 66.0 ± 11.1 (43-86) years and 25.6 ± 42.2 (5-195) days, respectively.

Procedure Isometric knee extensor muscle strength (maximum force) was measured twice (gravity eliminated) with a calibrated Ametek Accuforce II hand-held dynamometer. The dynamometer was applied just proximal to the malleoli on the anterior surface of the paretic leg while subjects were seated upright, their knees were in 90 0 of flexion, and they were manually stabilized by the examiner or an assistant. They were instructed to take 1-2 s to come to maximum effort and then to straighten their paretic knee as hard as possible until instructed to stop, after about 4 s more. Two knee extension efforts were separated by a rest of about 30 s. Knee extensor muscle force was expressed as a percentage of (normalized against) body weight. After the two measures of normalized knee extensor muscle strength were tested for reliability (intraclass correlation coefficient = 0.99), the mean of the two measures was calculated for use in all further analyses. Maximum weight-bearing was measured twice using calibrated Borg digital scales. Subjects stepped onto the middle of the two scales that were placed side to side on the floor. While looking straight ahead, they were instructed to slowly shift "as much weight as possible" onto their paretic lower extremity While keeping their paretic knee "slightly bent." They were guarded closely but not supported. They wore the same shoes that they wore during walking. The fOur subjects who wore an ankle foot orthosis during gait were allowed to wear it during the weight-bearing trials. The maximum weight noted on the digital display of the scale under the paretic lower extremity

CLINICAL PREDICTORS OF GAIT SPEED

....... 70 (/) <,

o

E

was recorded. After a 10-s rest, the process was repeated. Weight-bearing measurements on the paretic side were converted to a percentage of (normalized against) body weight. After the two measures of normalized paretic side weight-bearing were tested for reliability (intraclass correlation coefficient = 0.99), the mean of the two measures was calculated for use in all further analyses. Gait speed, which has been shown previously to be reliable (6,8),was measured only once. Subjects were asked to walk attheir most comfortable speed for 10 m while using their customary devices (cane, 17 subjects; orthosis, 4 subjects). The time required to traverse the last 8.0 m was determined with a digital stopwatch. Speed was calculated by dividing the distance walked (8.0 m) by the time (in seconds) required to walk it.

o

o

o

'-' 60

oUJ

o

UJ

fu

50

f-

< <:) 40

o o

30

Statistical Analysis

•

o

All statistical analysis was performed using the SYSTAT computer program (16). Pearson product moment correlations were used to determine zeroorder correlations among maximum weight-bearing, paretic knee extension strength, and comfortable gait speed. Regression analysis, both for linear and curvilinear models, was used to describe further the relationship between the measured variables.

20

10

o

,

,

i

,

i

20

40

60

80

100

PERCENT Figure 1. Scatterplot demonstrates the relationship between two

independent variables [paretic knee extension strength (force) and maximum weight-bearing through the paretic lower extremity] and adependent variable (gait speed) among 20 hemiparetic stroke patients. Each independent variable was normalized against (divided by) body weight to yield a percentage score.

Results Figure 1 illustrates the normalized paretic lower extremity weight-bearing and knee extension strength measurements associated with gait speed measure-

100

•

,...

•

~

• ••

..., 80

o Z

Figure 2. Scatterplot demonstrates therelation-

ship between paretic knee extension strength (force) and maximum weight-bearing through theparetic lower extremity among 20 hemiparetic stroke patients. Each variable was normalized against (divided by) body weight to yield a percentage score.

•• .._..!-.....-----•

~

~ 60

f-

:c

•

o

Ui

3: 40 :E ::> :E

X

-c

20

:E

o o

i

10

20

30

i

40

i

50

60

i

70

MAXIMUM KNEE EXTENSION FORCE (llb)

] STROKE CEREBROYASC DIS, YOLo 1, NO.2, 1991

67

R. W. BOHANNON

\0

0\ N

~

'8

\D

1:1::

..e

.~ lI)

~

..c

I

~

:J U

.~

II') ~

C

c::i

c::i

c::i

co

co

\0

0

~

)(

)(

..,.

'"50

I

...

\0

"l

'U

~

-s

~

E ~

'8

..s::

..

II> ;:

'" ClI

~

...

e

.,g 5 ..!:!

~

NX

~

...

+ + '1< ...

N

X C~M"'O\N

NCN~Ltiq

~c:io\CIf)C I II + II + II >. >. >.

co

a

0" lJJ

'':5 ;:

0\ N

If) c:i

e-,

(::! c:i

'U

~

~ '1:

II')

N

j:!:

"":

M

C

c:i

~

N

......

c:i

c:i

\D

~

'J::

t ~

...f"!

X

N

~

&ri \D I II >.

X

N

If)

co c:i

+ ...-0...

\D

,..;

+

\D

II e-,

+ 0\

\Ci

\D

Discussion

II >.

tlO

I..;

~

-;c .....

-

"ti

~ '.:::

c

ClI "'0 C ClI

....

.~

oS

0..

V)

41

C

0\

~

:0 ~

M

c:i

...

E

....

c:i

><

~ ;:s

.~

0-

c:i

..,.

~

'J::

C

~

II')

~

o

II')

c o

lJJ

g

co

1:1::

ClJ

-

"'0 C

e :0 ·2

.~ -- ~-- ~--

ClJ ClI C C ..c"'OClIoI::ClI..c ..!. .:: l: .~ J:: .~ ..c v, ell ClI II) ClI tlO.!:!C~C~

e 'C;;0 ~.c;;0

'- -

~

s-,

e ~E 2 E] : J C x ... x ... 5 0 tClI ~0 tClI ~0 ·x'?ft.

~""':z""'~"'"

~

til)

:B "EQl

~ ~ Ql

c

~ ~

..... ..... 13

..~ III

t3

13ClI

c C. ell

'n;

ments. All of the patients, who could ambulate with no more than contact guarding, were able to weightbear at least 44.4% of their total body weight through their paretic lower extremity. The mean (± SO) of the weight-bearing measurements through the paretic lower extremity was 83.4% (17.9%) of total body weight. The mean (±SO) of normalized knee extension strength measurements was 20.4% (13.3%). The range of normalized knee extension strength values was 0-58.2%. The association of the normalized paretic lower extremity knee extension strength measurements with the normalized paretic lower extremity weight-bearing measurements are illustrated in Fig. 2. Table 1 presents the results of the correlational and regression analysis. Both normalized weight-bearing and knee extension strength of the paretic side were correlated significantly (p < 0.01) with comfortable gait speed. Maximum weight-bearing, however, explained a greater percentage of the variance in gait speed (linear model, 52.9%; curvilinear model, 68.9%) than knee extension strength (linear model, 45.2%; curvilinear model, 45.6%). The variance in gait speed explained by the curvilinear model was substantially higher than the variance in gait speed explained by the linear model for weight-bearing measurements but not for strength measurements.

'e"'-.-

~~ , 01::

III u

.!::O :c

~ ~

e 0.. :J C

.5 ~

~~

::E

68 J STROKE CEREBROVASC DIS, VOL 1, NO.2, 1991

The demonstration of positive and Significant relationships among maximum weight-bearing through the paretic lower extremity, paretic knee extension strength, and comfortable gait speed among the tested patients was consistent with prestudy hypotheses. Although significant relationships between paretic knee extension strength and Comfortable gait speed have been reported before for patients with stroke (3-6), this is the first report of significant relationships between maximum weight_ bearing through the paretic lower extremity and Com_ fortable gait speed and between maximum weight_ bearing through the paretic lower extremity and paretic lower extremity strength. The finding of a significant relationship between maximum weight_ bearing through the paretic lower extremity and comfortable gait speed supports the validity of max_ imum weight-bearing as a measurement and as a potential therapeutic goal for use with patients With stroke. The range of the weight-bearing measure.. ments and the curvilinear nature of the relationship between maximum weight-bearing and gait speed in this study suggest that: (a) there may be a threshold of

CLINICAL PREDICTORS OF GAIT SPEED

maximum weight-bearing (approximately 40% of body weight) that must be surpassed before gait with no more than contact guarding can be accomplished by patients with stroke and (b) beyond some critical level (approximately 95% of body weight) further increases in maximum weight-bearing may not be particularly influential on gait speed. The correlation of maximum weight-bearing with comfortable gait speed in this study (r = 0.727) was higher than the correlation of weight-bearing asymmetry with comfortable gait speed (r = -0.324) reported in a previous study (9). This finding suggests that maximum weight-bearing through the paretic lower extremity may have greater value than weight-bearing asymmetry (during comfortable stance) as a predictor of gait speed in patients with stroke. The magnitude of the correlation between maximum weight-bearing and comfortable gait speed (R = 0.830) was substantially higher than the correlation between paretic knee extension strength and comfortable gait speed (R = 0.675). Thus, maximum weight-bearing is a superior predictor of gait speed. This is not surprising because, as was indicated earlier, maximum weight-bearing with the paretic knee slightly flexed should reflect the ability to use multiple lower extremity muscle groups (rather than one muscle group) to create an extensor moment and prevent collapse during stance. That paretic knee extensor muscle strength was correlated significantly (R == 0.700) with maximum weight-bearing suggests that paretic knee extension strength may determine a portion (49%) of the patients' ability to maximally weight-bear through the same lower extremity. Thus, the correlation provides further evidence of the validity of knee extensor muscle strength measurements in patients with stroke. Both maximum weight-bearing and knee extension strength are objective measures that are easy to obtain. The equipment required is portable and relatively inexpensive. Both measurements can be obtained from patients with stroke who are encountered in a variety of rehabilitation settings (including on the ward and at home). The measurements, therefore, can be advocated for predicting capacity and documenting progress in patients with stroke.

References 1. Winter DA. Energy generation and absorption at the

ankle and knee during fast,natural,and slowcadences. Clin Orthop Rei Res 1983;175:147-54. 2. Winter DA. Overall principle of lower limb support during stance phase of gait. 1Biomech 1980;13:923-7. 3. Hamrin E, EklundG,HillgrenAK. etal. Musclestrength and balance in post-stroke patients. UpsJMedSci1982; 87:11-26.

4. Bohannon RW. Selected determinants of ambulatory capacityin patients with hemiplegia. ClinRehabil1989; 3:47-53.

5. Nakamura R, WatanabeS,Handa T,et al. The relationship between walking speed and muscle strength for knee extension in hemiparetic stroke patients: a follow-up study. Tohoku J Exp Med 1988;144:111-3. 6. Bohannon RW, Andrews AW. Correlation of knee extensor muscle torque and spasticity with gait speed in patients with stroke. Arch Phys Med Rehab 1990;71: 240-3. 7. Bohannon RW.

Electromyographic activity of the quadriceps femoris muscles during four activities in stroke patients. Int 1Rehabil Res 1990;13:80-2. 8. Knutsson E, Gransberg L, Martensson A. Facilitation and inhibition of maximal voluntary contractions by the activationof musclestretch reflexesin patients with spasticparesis. EIectroencephalogr Clin Neurophysiol1988; 70:37P-8P.

9. Bohannon RW. Gait performance of herniparetic stroke patients: selected variables. Arch Phys Med RehabiI1987;68:777-81. 10. Bohannon RW, Larkin PA. Lower extremity weightbearing under various standing conditions in independently ambulatory patients with hemiparesis. Phys Ther 1985;65:1323-5. 11. CaldwellC,MacDonaldD,MacNeilK,et a1. Symmetry

of weight distribution in normals and stroke patients using digital scales. Physiother Pract 1986;2:109-16. 12. Mizrahi J, Solzi P, Ring H, et al. Postural stability in stroke patients: vectoral expression of asymmetry, sway activity, and relative sequence of reactive forces. Med Bioi Eng Comput 1989;27:181-90. 13. Winstein q, Gardner ER, McNealDR, et al. Standing balance training: effect on balance and locomotion in hemiparetic adults. Arch Phys Med Rehabil 1989; 70:755-62. 14. Cohen JJ, Sveen JD,WalkerJM,et al. Establishingcri-

teria for community ambulation. Top Geriatr Rehabil 1987;3:71-7. 15. Robinett CS,Vondran MA. Functional ambulation ve-

locity and distance requirements in rural and urban communities. Phys Ther 1988;68:1371-3. 16. Wilkinson1.. SYSTAT: thesystemforstatistics. Evanston, IL: SYSTAT Inc., 1987.

1STROKE CEREBROVASC DIS, VOL. 1, NO.2, 1991 69